CAUSESANDCONSEQUENCESOFHEALTHCAREEFFICIENCY:OVERVIEW U.S.healthcarespendingisatanall-timehigh,yetlifeexpectancyisstagnant,raisingconcernsthat projectedspendinggrowthwon?tleadtocommensuratehealthgains.Thisisaproblemespeciallypertinentto olderadults,withahighburdenofillnessandintenseuseofhealthcare.InthisP01renewalapplication,we proposetocontinueourlong-standingworkidentifyingefficiencyandinefficiencyinU.S.healthcareby applyingsophisticatedempiricalmethodstomorethan1billionperson-yearsofhealthdata.Weaddress3 topicsintegraltohealthcaredeliveryefficiency:First,ourexplorationofchallengesinclinicaldecisionmaking isexpectedtoidentifysystematicunderuseofeffectivecareandoveruseofineffectivecare,andprovide evidencewhereclinicaltrialsarelacking.Second,wewillexaminetheroleofthehealthcaredelivery environment?howdopatient-sharingnetworks,paymentmodels,andregulationsaffectpatienthealth?Third, throughourmulti-payerdata,weseektounderstandthelimitationsofpolicyanalysisarisingfromafocuson justMedicareorprivate(commercial)data.Forexample,weseektomeasurehowhighcommercialpayment ratesaffectaccessforMedicareorMedicaidenrollees.Allprojectsaimtomeasureandimprovequalityofcare forolder,vulnerablepopulations,includingpeoplewithAlzheimer?sDiseaseandrelateddementia(ADRD). Wepropose3coresand5projects.CoreAprovidesadministrativesupport,CoreBcoordinatesdata;? andCoreC(Methods)developsnewapproachestonetworkanalysisusedinall5projects.InProject1, ?CorrelatesandConsequencesofMakinganAlzheimer?sDiseaseClinicalDiagnosis,?weusethe100% MedicarefilesandclinicallyrichsurveydatatodocumentvariationinADRDdiagnosis,andtestwhetherearly diagnosisis,onnet,beneficialtopatients.Project2,?TheCausesandConsequencesofRiskyPrescribing,? usesMedicarePartDprescriptiondatatostudyadverseoutcomesassociatedwithindividualdrugsanddrug combinationsincludingopioidanalgesics,benzodiazepines,andsedativehypnotics.Weconsiderforces influencinghigh-riskprescribing,suchasshared-patientnetworksandlegalrestrictions.Project3,?Identifying EfficientHealthCareProviders:EvidencefromHospitalClosuresandRegistryData?usesperipheralvascular disease(PAD)registrydata,andclaimsdata,toestimatetherelativeexpertiseofcommunityhospitals.This projectwillvalidatemethodswithnaturalexperiments(hospitalclosing)andanongoingrandomizedtrial. Project4,?CausesandConsequencesofVariationinPublicandPrivatePaymentRates,?hypothesizesthat highcommercialreimbursementratescannegativelyaffectaccesstocareforMedicareandMedicaidpatients. Finally,Project5,?PhysicianCognition,InpatientAdvance-CarePlanning,andOutcomesforSeriouslyIllOlder Adults,?usesacombinationofobservationalandexperimental(randomized-trial)researchin250community hospitalsacrosstheU.S.toimprovethequalityofinpatientphysiciandecision-making.